Expired60% confidenceFactExact time
Mainstream open-source code models include Meta's Code Llama series (parameter scales ranging from 7B to 70B), DeepSeek Coder, and Codestral from Mistral.
2
Sources
60%
Confidence
Medium-term (~90 days)
Relevance
7/7/2026
First Seen
Valid until: 10/5/2026(expired)
Sources
Codex配置实战：国内低成本接入GPT-5.5完整教程
bilibilidream黎6/7/2026
Related Claims
VerifiedCodeLlama was trained by Meta through continued pre-training and instruction fine-tuning of Llama 2 in the code domain, offered in multiple parameter scales including 7B, 13B, and 34B.79% similarUnverified主流可本地运行的编码友好模型包括Llama 3系列、阿里的Qwen2.5-Coder、DeepSeek-Coder、Mistral以及微软的Phi系列78% similarUnverified当前最强开源代码模型如DeepSeek Coder V2、CodeQwen参数规模通常为7B-70B级别，与Claude等闭源前沿模型仍存在明显能力差距75% similarUnverifiedCodeLlama is a code-specialized model produced by Meta.75% similarUnverifiedCode large models such as DeepSeek Coder, CodeLlama, and StarCoder all follow a technical path similar to that of Codex72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/147295API
curl https://kongchang.com/api/v1/knowledge/claims/147295MCP
get_claim(id=147295)